One question and a followup comment:
In the intro to the BayCHI meeting, it talks a bit about developments in social and personalized search. But I don’t quite understand the claim:
how do recommendations based on personal tastes, tags, popularity, etc “go beyond the classic search paradigm of relevance”?

Relevance is defined as something that fills my information need, right? Tastes, tags, popularity, etc are just features that help one assess relevance. But they’re not somehow “beyond” relevance.

The “classic search paradigm of relevance”, as it has been defined since the 1960s (take a look at the Cranfield experiments), is and has always been to find items that meet a user’s information need. If that information need is to find something you didn’t know existed (e.g. new music) as opposed to simply finding the home page for a company you know does exist, that doesn’t matter. In both cases, what you’re after is filling your information need, i.e. relevance.

Or is there some other understanding of this that I’m not getting? Wish I coulda made it last night, and asked folks there a bit more about this.